The long term purpose of this research is to gain understanding of the way mechanical variables are encoded in the discharge pattern of mechanoreceptors in the joint capsule of the knee. Two specific questions will be addressed, using different methodologies. One specific aim is to identify whether stress, strain, or strain energy density is best encoded by afferent neuron discharge. Single neurons will be recorded in vitro in a one dimensional strip of joint capsule, subjected to a wide range of dynamic stimuli. Mechanical and neuronal responses will all be measured, and information theoretic methods will be used to examine the encoding of mechanical variables by neuronal responses. Once a mechanical variable is identified as the input variable for mechanoreceptors, a second aim is to characterize the system input- output relationship between that variable and a pseudorandom white noise paradigm, and nonlinear systems analysis methods will be used to compute the set of specific relationships (kernels) which characterize the neuronal output in terms of the input variable. These studies will be done with both one dimensional and two dimensional loading paradigms.